研究目的为通过人工神经网络方法预测玻璃钢天线罩的电子性能.建立具有非线性逼近能力的径向基函数(RBF)神经网络,根据试验得到的不同厚度玻璃钢平板,不同入射角的透波率数据,对神经网络进行训练.按照给定的玻璃钢天线罩内外表面数据计算入射角范围和罩壁厚度,并对玻璃钢壳体进行电子性能预测.计算结果与试验数据十分近似,表明该方法预测精度高,训练速度快,为玻璃钢电子性能设计和分析提供了一种实用有效的方法.
The aim of the research is to predict the electrical performance of FRP radome via artificial neural network. First, the radial basis function (RBF)with strong capability of nonlinear approximation was constructed. Then, RBF was trained by experimental data include incidence angle, transmittance and thickness of FRP radome. The incidence angles and thickness were calculated when the external and inner surface data of FRP radome were given. At last, the electrical performance of FRP radome was predicted by the trained network. The re-sults proved the precision of prediction was precise, the train rate of RBF neural network was fast. The method was valuable for radome design and analysis.
参考文献
| [1] | 杜耀惟.天线罩电信设计方法[M].北京:国防工业出版社,1993 |
| [2] | 阎平凡;张长水.人工神经网络与模拟进化算法[M].北京:清华大学出版社,2000 |
| [3] | 于霖冲 .复合材料机载雷达罩机电性能设计技术的研究[J].沈阳航空工业学院学报,2000,17(03):42-43. |
| [4] | 葛建华,王迎军,郑裕东.人工神经网络在材料科学与加工中的应用[J].现代化工,2003(01):59-62. |
| [5] | 焦俊婷,于霖冲.基于ANN的复合材料变厚度壳体固化变形预测[J].玻璃钢/复合材料,2006(05):3-5,31. |
| [6] | 焦俊婷,于霖冲.基于RBF的玻璃钢天线罩逆向工程曲面重构[J].玻璃钢/复合材料,2007(01):6-8. |
- 下载量()
- 访问量()
- 您的评分：
-
10%
-
20%
-
30%
-
40%
-
50%